1. Field of the Invention
The present invention relates to a white paper detection method of an image processing apparatus.
2. Description of the Related Art
When a user transmits facsimile or makes a copy of a document by a digital multifunction peripheral, in some cases, a wrong face of the document is unintentionally set by the user or white paper is unintentionally included in the documents to be transmitted or copied. Under such circumstances, a technique referred to as “white paper detection” is generally used. Japanese Patent Application Laid-Open No. 11-284846 discusses a technique that realizes detection of white paper by dividing a scanned image into blocks and calculating a ratio of black pixels in each block.
Further, Japanese Patent Application Laid-Open No. 2002-165054 discusses a technique capable of handling a color document in addition to detection of a white paper document. The technique is useful in determining whether the document includes an object to be read by comparing the change in the number of black pixels which have been read for each line in the main scanning direction and a threshold value. More specifically, the black pixels in the document is read for each line, and if the amount of change of the number of black pixels for one line is smaller than the threshold value, the document to be scanned is determined that it is the white paper document that does not include an object.
In recent years, environmental issues have received widespread attention, and the use of environmentally-friendly recycled paper has increased in documents to be used in printing. In some cases, foreign substances in the material of the recycled paper can be visually recognized. Such recycled paper and poor-quality paper including especially high ratio of foreign substance is also used for printing. (In the description below, both the recycled paper and the poor-quality paper are collectively referred to as low-quality paper.)
In this manner, the use of low-quality paper as document paper has increased. From the viewpoint of white paper detection, it is desirable to be able to maintain the detection accuracy even when the low-quality paper is used. However, regarding conventional techniques, if the paper quality of the document is low and a great number of foreign substances are included in the paper, it is difficult to accurately detect the white paper. This is because, according to the technique discussed in Japanese Patent Application Laid-Open No. 11-284846, if the document includes a great number of foreign substances, it is difficult to distinguish between black pixels that indicate foreign substances and black pixels that indicate the document content. Further, regarding the technique discussed in Japanese Patent Application Laid-Open No. 2002-165054, if a great number of foreign substances are included in the paper and the amount of change of the black pixels in one line is increased, it is difficult to determine whether the document to be read is a document including print information or a white paper document.
Further, although the setting of a threshold value for white paper detection can be changed when low-quality paper is used, if the documents to be read are a mixture of low-quality paper and high-quality paper, since an optimum threshold value differs according to the documents, it is difficult to accurately perform the white paper detection for all the documents.
Further, although a pixel with a density lower than a certain density can be rounded to a white pixel, since the density of the pixel that corresponds to the document content cannot be distinguished from the density of the pixel of the foreign substance, it is difficult to round only the pixel of the foreign substance to a white pixel.